Thierry Faye, Agilent Technologies Jef Focant, University of Liege, Belgium

Transcription

1 Analysis of Dioxins and Dioxin-like PCBs in Feed and Food Thierry Faye, Agilent Technologies Jef Focant, University of Liege, Belgium

2 GC/MS/MS Analyzer for Analysis of Dioxins and Dioxin-like PCBs in Feed and Food Thierry Faye Market Development Manager Agilent Technologies

3 Introducing the Agilent GC/MS/MS Dioxin Analyzer Cutting edge technology enables fast & reliable Dioxin analysis at low levels: New standard for GC/MS/MS sensitivity with 7010 Triple Quadrupole EI source = An instrument with sensitivity of GC Sector MS. MultiMode Inlet (MMI) for effective cold split less injections and more. Developed from successful collaborations with leading Dioxin Labs in Europe: The Agilent platform already validated according to new regulations in Europe for both food and feed (EC 589/2014, 709/2014). Refer to L Homme/Focant s publication: Journal of Chromatography A, 1376 (2015) Custom reporting with complete calculations have been developed and automated in MassHunter. Ready for analysis: Pre-configured and pre-tested at our factory so installation in your lab is fast and efficient. The RTL advantage: guarantees the exact matching of our reference method on a new instrument. A service engineer runs a compete check out standard so validation can begin. Method never needs altering even when column maintenance is performed. Agilent Confidential 5/4/2015

4 Validation of GC/MS/MS confirmatory method for the European official control of levels of dioxins, furans, and dioxin-like PCBs in foodstuffs L Homme B., Focant J.-F. Organic and Biological Anal. Chem. University of Liege, Belgium

5 Yusho Yu-Cheng

6 95% Exposure by Food Consumption Food and Feed control

7 Dioxin EU Regulation for food-feed started with the Belgian Dioxin Crisis in 1999

8 EU Commission Documents Council Regulation 2001/102/EC Council Regulation (EC) No 2375/2001 Commission Directive 2002/69/EC Commission Directive 2002/70/EC Commission Directive 2006/13/EC Commission Recommendation 2006/88/EC Commission Regulation (EC) No 199/2006 Commission Regulation (EC) No 1881/2006 Commission Regulation (EC) No 1883/2006 Commission Regulation (EC) No 152/2009 Commission Regulation (EU) No 1259/2011 Commission Recommendation 2011/516/EU Commission Regulation (EC) No 252/2012 Commission Regulation (EC) No 277/2012 Commission Regulation (EC) No 278/2012 Commission Regulation (EC) No 589/2014 Commission Regulation (EC) No 709/2014

9 EU Commission Strategy (food-feed) Continuous monitoring Maximum-Action-(Target) level strategy Screening-Confirmatory approach RASFF (high capacities) Based on state-of-the-art methods

10 Analytical Methods?

11 Evolutive Guidelines PBMS OHC 50 (2001) 53 OHC 50 (2001) 59 GC-IDHRMS vs CALUX ISO17025 LOQs (@1/5 th level of interest) Upper-Lowerbound < TEQ/g fat level PTs Recovery rates

12 Recent EN 16215

13 Xtraction Clean-up & Fractionation Analysis

14 Confirmatory Tool HRMS sector or MS/MS instruments Organohalogen compound 74 (2012), EU Regulation 252/2012 amended by 589/2014 (2 nd of June 2014) EU Regulation 152/2009 amended by 709/2014 (20 th of June 2014)

15 New EU Regulation

16 PCDD/Fs and DL-PCBs NDL-PCBs Criteria GC-MS/MS (589/2014) GC-MS/MS (589/2014) Detectable quantity Selectivity MRM tansitions Blank iloq LOQ Accuracy Reproducibility Control Recovery Measurement uncertainty -PCDD/F upper femtogram (10^-15g) -NO-PCB low picogram (10^-12g) NDL-PCB nanogram (10^-9g) -MO-PCB nanogram (10^-9g) -Chromatographic separation of 1,2,3,4,7,8-HxCDF and 1,2,3,6,7,8-HxCDF Relative RT ±0.25% IS vs analyte <25% valley peak to peak -Monitoring 2 specific precursors with -Monitoring at least 1 precursor ion and each specific product ion transition for all 2 product ions labeled and unlabeled analytes -Tolerance ratio ±20% if rel. intens. >50% -Relative ion intensities max ±15% Tolerance ratio ±25% if rel. intens % -Resolution MS quadrupoles = unit -Resolution MS quadrupoles = unit -Used for LOQ calculation -Used for LOQ calculation -Blank value <30% of maximul level ML -iloq calculated from lowest cali. point -lowest concentration point on cali. must give acceptable and consistent deviation to the average RRF -ditto -Average RRF calculated for all points -Deviation to average RRF <30% -LOQ caluclated from average blank level -LOQ < 1/5 of maximum level ML -ditto -Difference ub and lb levels <20%ML -Diff. ub and lb for sum ML <20% -Demonstrate performances at 0.5ML, ML, 2ML -Demonstrate performances at 0.5ML, ML, 2ML -Trueness (accuracy) ±20% -Trueness for sum ML ±30% -Within-lab reproducibility (RSD) <15% -Within-lab reproducibility (RSD) <20% -QC chart for blanks -QC chart for blanks -QC charts control sample -QC charts control sample -Individual internal std in range % -Individual internal std in range % -Out of range OK if contribu. to TEQ<10% -Out of range OK if contribu. to sum ind-pcb<10% -Expanded measurement uncertainty -Coverage factor = 2 (CL=95%) -Expanded measurement uncertainty -If separate determination of congeners, -Coverage factor = 2 (CL=95%) make sum of separate uncertainty for sum of PCDD/F and DL-PCBs From Regulation Full validation

17 Validation for vegetable oil (feed) MU Top-down approach Control Control chart Blanks subtraction Average blank Blanks control chart Agilent 7000C series Triple Quad GC/MS Reproducibility Recovery experiments, QCs Accuracy LOD/LOQ Recovery experiments Proficiency test Statistical determination Matrix related Selectivity, linearity MRM, Quant/qual transitions GC separation, Cali. curve

18 PTV Injections Dioxin fraction runtime = 42min MO-PCBs fraction runtime = 28min

19 Chromatographic Profile MO and NDL PCBs PCDDs, PCDFs, NO PCBs

20 Chromatographic Separation -> 310.9) QCG_9_162.D Smooth Selectivity x10 2 +EI MRM CID@35.0 ( > 310.9) QCG_9_162.D 3 * * > 310.9) QCG_9_22.D Smooth x10 2 +EI MRM CID@35.0 ( > 310.9) QCG_9_22.D 1.2 * * > 310.9) QCG_9_9.D Smooth 0.4 x10 2 +EI MRM CID@35.0 ( > 310.9) QCG_9_9.D * * x10 2 +EI MRM CID@35.0 ( > 310.9) QCG_9-122.D -> 310.9) QCG_9-122.D Smooth 1.5 * * * * x10 2 +EI MRM CID@35.0 ( > 310.9) QCG_9-62.D * * > 310.9) QCG_9-62.D Smooth x10 2 +EI MRM CID@35.0 ( > 310.9) QCG_9-67.D 3 * * > 310.9) QCG_9-67.D Smooth * * Counts vs. Acquisition Time (min) me (min) 25% valley separation HxCDF

21 Tandem in-space MS

22 MRM transition Ratio Quant/Qual transition ratio EU Reg 709/2014 says: PCDD/Fs, DL-PCBs: 2 specific precursors with each specific product ions 2378 TCDD: > TCDD: > PCB 189: > PCB 189: > Average ratio = NDL-PCBs: at least 1 precursor and 2 product ions Average ratio = Tolerance ± 15% R(quad)=unit ± 20% R(quad)=unit

23 In MassHunter Quant/qual MRM transitions SANCO says: PCDD/Fs, DL-PCBs: 2 specific precursors with each specific product ions 2378 TCDD: > TCDD: > Average ratio = NDL-PCBs: at least 1 precursor and 2 product ions Tolerance: 95.1 ± 15% Tolerance ± 15% R(quad)=unit PCB 189: > PCB 189: > Average ratio = ± 20% R(quad)=unit

24 Proper Estimation of LOQs LOD/LOQ Organohalogen compound 74 (2012), iloq = 10*stdev (8 replicate injections cali point) LOQ = blank mean + 6*stdev (12 injections blank)

25 Validation for vegetable oil (feed) ilod/iloq Acceptable and consistent deviation to the average RRF < 15% Av RRF 1-6 = R 2 (linear fit) = > Av RRF 1 = Deviation = -1.9% < 30%

26 Validation for vegetable oil (feed) ilod/iloq Calibration curve (lowest level) 8 replicates LOQ matrix Procedure blanks 10*stdev 10 independent injections Average + 6*stdev.

27 Validation for vegetable oil (feed) (i)loq Lowest cali point Blanks Compound Avg Conc. pg/µl Std. Dev. ilod pg/µl iloq pg/µl in blank? LOQ MS/MS pg/g LOQ HRMS pg/g PCB 77 0,2917 0,0037 0,011 0,037 yes 49,66 64,59 PCB 81 0,2935 0,003 0,009 0,030 yes 3,67 9, TCDD 0,0152 0,0018 0,005 0,018 no 0,02 0, TCDF 0,0162 0,001 0,003 0,010 yes 0,10 0,12 PCB 126 0,2906 0,0077 0,023 0,077 yes 1,37 1, PeCDF 0,0154 0,0021 0,006 0,021 yes 0,08 0, PeCDD 0,0169 0,0029 0,009 0,029 no 0,01 0,06 PCB 169 0,3062 0,0071 0,021 0,071 no 0,02 0, PeCDF 0,0142 0,0022 0,007 0,022 yes 0,56 0, HxCDF 0,0164 0,0016 0,005 0,016 yes 0,07 0, HxCDF 0,0169 0,0009 0,003 0,009 yes 0,06 0, HxCDF 0,0166 0,0007 0,002 0,007 yes 0,08 0, HxCDF 0,0156 0,002 0,006 0,020 yes 0,18 0, HxCDD 0,0774 0,0061 0,018 0,062 yes 0,06 0, HxCDD 0,017 0,0022 0,007 0,022 yes 0,02 0, HxCDD 0,0426 0,0032 0,010 0,032 yes 0,14 0,08 OCDD 3,5201 0,0465 0,140 0,465 yes 1,99 2,02 OCDF 0,0158 0,0027 0,008 0,027 yes 0,60 0,77

28 Accuracy & Reproducibility Spiked materials at 0.5 ML, ML, 2 ML (6 series, 3 days) Bias < 20% Within lab reproducibility < 15%

29 Accuracy & Reproducibility Proficiency test (PT) Z score = 0.80 Z score = 0.59 Wim Traag, Rikilt, NL

30 Within Lab Reproducibility QC pork fat

31 Sum TEQ PCDD/F 0,175 ± 0,032 Sum TEQ PCBs 0,146 ± 0,035 MassHunter Report Generator D:\MassHunter\GCMS\1\data\Develop PCDD-F-NOPCB\QuantResults\Laberca-PCDD-F-NO.batch.bin Batch Data Path Analysis Time :33 Analyst admin Report Time :30 Reporter admin Last Calib Update :56 Batch Laberca-PCDD-F-NO.batch.bin T16:56: :00 12T10:33: :0 12T15:30: :00 Sample Name _diox-1 Type Sample Vial 62 Vol. [µl] 5 Comment Diox Compound RT [min] Conc [ng/ml] TEQ Conc [ng/ml] LOQ Upper Bound [ng/ml] Medium Bound [ng/ml] Lower Bound [ng/ml] WHO-TEF TCDD 20,94 <LOQ <LOQ 0,0200 0, , , PeCDD 24,27 <LOQ <LOQ 0,0300 0, , , HxCDD 28,02 0,0376 0, ,0200 0, , , , HxCDD 28,15 <LOQ <LOQ 0,1400 0, , , , HxCDD 28,51 <LOQ <LOQ 0,0600 0, , , , HpCDD 33,08 <LOQ <LOQ 0,4400 0, , , ,01 OCDD 39,46 4,9895 0, ,9900 0, , , , TCDF 19,90 <LOQ <LOQ 0,1000 0, , , , PeCDF 23,26 <LOQ <LOQ 0,5600 0, , , , PeCDF 24,08 <LOQ <LOQ 0,0800 0, , , , HxCDF 27,05 <LOQ <LOQ 0,0700 0, , , , HxCDF 27,19 <LOQ <LOQ 0,0600 0, , , , HxCDF 29,06 <LOQ <LOQ 0,1800 0, , , , HxCDF 27,84 <LOQ <LOQ 0,0800 0, , , , HpCDF 31,29 <LOQ <LOQ 0,5500 0, , , , HpCDF 34,01 <LOQ <LOQ 0,0200 0, , , ,01 OCDF 39,89 <LOQ <LOQ 0,6000 0, , , ,0003 PCB 77 18,01 <LOQ <LOQ 49,6600 0, , , ,0001 PCB 81 18,01 6,7691 0, ,6700 0, , , ,0003 PCB ,04 <LOQ <LOQ 1,3700 0, , , ,1 PCB ,22 <LOQ <LOQ 0,0700 0, , , ,03 Total TEQ 0, , ,00729 Sum TEQ PCDD/F 0, , ,00526 Sum TEQ PCBs 0, , ,00203

32 Take Home Message PTV-GC/MS/MS accepted as a confirmatory tool under EU Regs Full validation on challenging matrix MS/MS, but still dioxin analyses MS/MS & sectors to be properly perceived